OpenAI Decisions API lacks trustworthy confidence metrics
Problem
Users are confused by the API's confidence description, especially the frequent use of vague terms like 'usually', making it hard to assess output reliability. Moreover, commenters note that what is presented is actually the Luna model, not the Decisions API, further obscuring usage guidance.
Pain points
- Ambiguous confidence description in documentation, lacking clear trustworthy metrics
Questions to verify
- Has the documentation provided a clear definition of confidence?
- Do developers need more examples on how to use the Decisions API instead of Luna?
Evidence
quote
The problem statement includes the word “usually” a few times.
Disclaimer: this card is distilled by AI from community discussions as “pain point evidence”. It is a starting point for need analysis and validation, not a “validated market need”. AI has analytical limits; please open the original discussion above to manually verify before key decisions.